Sistotrema luteoviride sp. nov. (Cantharellales, Basidiomycota) from Finland

A new Sistotrema species from Northern Finland, S. luteoviride is described and illustrated. The two hitherto known collections derive from Finnish Lapland and both grew on corticated Juniperus communis . The spores are very similar to those of S. citriforme , which however is a simple septate species and differs clearly by its ITS sequence.


INTRODUCTION
Sistotrema Fr. is a comparatively large genus (Index Fungorum 2013) typified by the stipitate species S. confluens Fr. Despite the morphology of the type, all other species presently referred to Sistotrema have effused basidiocarps with a smooth, hydnoid or poroid hymenophore. The type species together with a few poroid or hydnoid species probably all have an ectomycorrhizal habit (Nilsson et al. 2006;Münzenberger et al. 2012) while the majority of species seem to be saprophytes. According to Nilsson et al. (2006) the genus is non-monophyletic, and most likely the species outside the core group around the type must be distributed over several genera (Larsson 2007).
During a survey of corticioid species dwelling on Juniperus communis L. in Estonia (Sell, Kotiranta 2011) and Finland, two collections of an unknown species with morphological similarities to Sistotrema citriforme (M.P. Christ.) K.H. Larss. & Hjortstam were discovered in Finnish Lapland. Molecular phylogenetic analyses confirmed its affinity to Sistotrema and its close relationship to S. citriforme. The species is here characterized and described based on morphological and molecular information.

MATERIALS AND METHODS
Thirty spores per specimen were measured in Cotton Blue (CB) or Melzer´s reagent (IKI). CB+ means that the walls of the cells are stained by Cotton Blue and CB-, that they are not stained, and IKI-that there is no reaction to Melzer´s reagent.
The following abbreviations are used: L* = mean spore length, W* = mean spore width, Q = range of the variation in L/W ratio, Q* = quotient of the mean spore length (L/W). None of the measurements derive from a spore print.
Biological province of the collecting site in Finland is indicated according to the Finnish national uniform grid system (27°E), as applied to biological material by Heikinheimo and Raatikainen (1981). Nine sequences from the nuclear ribosomal ITS and LSU regions are newly published for this study. Data on vouchers and on sequences downloaded from GenBank and included in phylogenetic analyses are provided in Table 1. DNA was extracted from dried specimens (Tab. 1) using DNeasy plant mini kit (Qiagen), following manufacturer's recommendations. PCR reactions were carried out using Ready-To-Go™ PCR beads (Amersham Pharmacia Biotech). Primers used to amplify the complete ITS region and the 5´end of the LSU region were ITS1F and ITS4B (Gardes, Bruns 1993), and LR0R and LR7 (Hopple, Vilgalys 1999) respectively. Amplified products were purified using Qiaquick spin columns (Qiagen). Primers used for sequencing were ITS3, ITS4 (White et al. 1990), LR5, LR3R (Hopple, Vilgalys 1999), and CTB6 (http://plantbio.berkeley.edu/bruns/). Sequencing was done by Macrogen (South Korea). Sequences were edited and assembled using Sequencher 3.1 (Gene Codes, Ann Arbor). A dataset with 12 sequences representing Sistotrema, Membranomyces Jülich, and Clavulina J. Schröt. was compiled. Protodontia piceicola (Bourdot) G.W. Martin was added as outgroup. The dataset had 2378 nucleotide positions when the 5.8 and LSU regions had been manually aligned. ITS1 and 2 were left unaligned and not included in the analysis. The analysed matrix had 1397 characters of which 115 were parsimony informative and 183 variable but uninformative. Maximum parsimony was performed using PAUP* 4.1b10 (Swofford). All transformations were considered unordered and equally weighted. Gaps were treated as a fifth character. Heuristic searches used 500 random taxon addition replicates and TBR branch swapping with other options using the program's default settings. Relative robustness of clades was estimated through bootstrap analysis using PAUP*. Settings used were 500 bootstrap replicates with 100 random addition sequences per replicate, TBR branch swapping, and otherwise default settings.

RESULTS
The phylogenetic analysis generated 4 most parsimonious trees, one of which is shown as a phylogram in Figure 3. This tree had a consistency index 0.7137. The new species clusters with Sistotrema citriforme with high bootstrap support (98%). They form together with S. pistilliferum Hauerslev, Membranomyces spurius (Bourdot) Jülich, and two Clavulina species a moderately supported clade (79%). The ingroup is not resolved as monophyletic and the paraphyletic nature of Sistotrema is again indicated (Nilsson et al 2006). However, the restricted dataset used in this study was not designed to reveal any new information regarding the phylogeny of Sistotrema.
The ITS sequences of Sistotrema citriforme and S. luteoviride described below differ by 70-73% depending on alignment method and including the conservative 5.8S gene.
Even if there are hitherto only two specimens of S. luteoviride, we believe that it is not an especially rare species in the subarctic areas of the Northern Hemisphere. The habitats, river-or brook-side forests, nutritionally poor Pinus sylvestris dominated forests, and alpine heath vegetation, are very common in the north. The substrate in both collections was a corticated Juniperus communis, but if we believe that the species is a symbiont common juniper is unlikely to be the host. Rather the juniper bushes happened to offer a suitable substrate where S. luteoviride could develop its resupinate, effused basidiocarps.